SEEDGuard.AI An Open Source Library for Building High-Quality Data for Trustworthy AI for SE

What is SEEDGuard.AI? Revolutionizing Trustworthy AI for Software Engineering!

Welcome to SEEDGuard.AI, an open-source initiative developed collaboratively by a team at North Carolina State University and global experts to redefine AI in software engineering by improving data quality. SEEDGuard.AI focuses on addressing data quality issues in the AI for SE/Code domain. SEEDGuard, short for Software EnginEEring Data Guard, reflects our commitment to safeguarding the integrity and reliability of software engineering datasets.

In data-driven software engineering, the significance of high-quality datasets is crucial. SEEDGuard.AI acknowledges the essential connection between dataset quality and the success of data-driven software engineering. Our vision is to create a data-centric library tailored for both researchers and practitioners, especially those working on Large Language Models.

Whether you're a professional or new to the open source community, SEEDGuard.AI invites contributions from all. Together, we aim to elevate the standards and quality of data for the benefit of the entire software engineering community.

Check Our How to contribute for more information!


Defense software engineering data from poisoning attack.


Quality assessment for software engineering data.


Augmenting software engineering data.